WILL DYER (@W2Dyer) explains how his unique match ratings system can be used to predict results by regular punters.
Predicting Results With A Match Ratings System
My latest foray into betting analysis has come by the way of ‘Match Ratings’. Many successful and professional punters use systems like this to predict match outcomes and so I thought I’d give it a go. I’ve kept records since the start of the 2015/16 Premier League season.
The last few weeks have seen the system start to bear some pretty splendid fruit. After 17 gameweeks my Match Ratings are showing an LSP (Level Stakes Profit) of +37 points. Where 1 point is equivalent to 1 stake per-match – this means, if you had £10 on each and every one of my 170 Premier League predictions you’d be £370 in profit; which, given the incredibly ‘unpredictable’ season we are having, is pretty good going.
The Match Ratings themselves are pretty simple. I give each team marks out of 10 for their match performance. By match performance I don’t at all mean how good the final score was for said team, instead I mean how well they played, in my opinion. They could play well and be extremely unlucky and lose 3-0, it happens; and quite regularly too.
Recent form can be deceiving. Results do not always tell the whole story – pretty often they are not a fair reflection of the game – and I find that teams that are playing better than their recent results suggest will regularly win games that you couldn’t have been able to predict without some sort of record.
I have found that you don’t need to watch every single game live to give a team marks out of 10. In an ideal world you would but that’s unrealistic. As a fan of the Premier League in general, as I’m sure most of you reading this are, then simply watching Match of the Day and a couple of live games each week and reading the newspapers or articles online is more than enough to keep records effectively and make fair judgement.
Judging performances fairly is a lot easier to do when you are a neutral. As a Swindon fan I don’t have to worry too much about my team being in the Premier League any time soon so I have no allegiances to any side. Of course there are always teams you prefer or wish to do well more than others but I try to be as impartial as possible.
How The Strategy Works
So I score each team out of 10 each week and total those scores up. This weekend coming is gameweek 18 so the current maximum points tally is 170 from the previous 17 gameweeks. I then compare the ratings of the teams playing each other and come up with a prediction based upon the difference between their total scores.
Earlier in the season I was using 10% as my differential but I’ve reduced that to 5% after 10 gameweeks because it was creating too many draw predictions. So using this 5% differential I look at each side’s totals for that week and compare that to their opposition.
For example, If Team A has more than 5% of the maximum than Team B then I select them to win the game. So the maximum for this week being 170 means my 5% differential is 8 (rounded down from 8.5). If Team A and Team B are separated by 5% or less then my prediction is draw.
Example A: Aston Villa (102/170) v West Ham United (114/170).
Example B: Tottenham Hotspur (119/170) v Norwich City (114/170)
In Example A, West Ham have 12 points more in my Match Ratings than Aston Villa so that is more than the 5% differential of 8, which means that I am selecting West Ham to win.
In Example B, the teams are separated by just 5 points which is less than the 5% differential, so I am predicting the match will finish as a draw.
This is another area where I think this strategy could excel and already has for me. People do not bet on the draw enough, it’s a well-known phenomenon, and I think being persuaded to by the match ratings pointing to a close game is a good thing.
It could also help you avoid backing weak or under-priced favourites. Since changing the differential to 5% in Week 11 my Match Ratings have correctly predicted 11 of 25 draws, which has given a profit figure of +18.4 LSP – almost half of the total profit that the system has so far generated.
How To Use It In Your Betting Strategy
It would be unfair to judge your predictions and their accuracy too crudely. Just because Watford are having a good season and playing well doesn't mean you have to back them to win at the Emirates, for example.
Rather I like to choose my favourite predictions each week and back those (which is usually in the region of four or five matches) Match Ratings of which compliment my own perception or gut feeling on the game and all other factors well.
That being said, my system is showing a great profit figure overall when including every game in the Level Stakes Profit (LSP) total. Should you develop your own system and see it achieve similar success then you may choose to back every prediction. After all, a lot of the profit from my season so far has come from predictions that look highly unlikely to come off.
The key to all of this is that it is based on opinion and if you are too biased in your judgement it probably won’t give you good results in your betting. But I feel that it helps clear the mind to get things down on paper or spreadsheets.
You may well look at a fixture and decide that your Match Ratings have no bearing on how you think the game will pan out whatsoever and that’s fine but, all in all, I think it can have a good part to play in your betting research alongside checking the form and team news. And it only takes a couple of minutes to do each week…
We're now offering all players a chance to win £100 cash from us.
Just sign-up with a bookmaker via our site, deposit then email us at firstname.lastname@example.org.
We'll offer you a cash prize straight away but you can gamble that up to £100 by picking winners.